Explosion Risk Analysis (ERA) is typically performed to evaluate the explosion Design Accidental Loads (DALs) for the topside structures and facilities on offshore platforms. The total number of Vapor Cloud Explosion (VCE) scenarios in the real world is innumerable. To deal with the entire range of possible scenarios, the ERA is typically performed in a probabilistic manner, where the actual scenarios are replaced by a certain number of representative scenarios. A representative scenario is defined by several independent variables such as gas cloud volume, position and shape, and each of them should be probabilistically determined by using the concept of a flammable gas cloud frequency distribution. However, since an actual gas cloud is an object that extends into the 3D space, it is difficult to determine the shape of the actual gas cloud (or the fuel concentration distribution) with a particular scalar variable. In the existing ERA approach, therefore, the shape cannot be determined by the flammable gas cloud frequency distribution but is conservatively assumed to be a rectangular or other type of simple 3D geometry. Beyond the accuracy of this assumption, consistently determining the aspect ratio or other geometric parameters of the applied 3D geometry is a big issue in practice. Currently, for a given gas cloud volume, there is little guidance on how to determine the geometric parameters of the applied 3D geometry, and the relevant data is eventually left to the engineering judgment and experience. Therefore, in most cases, ERA results may vary from engineer to engineer, rather than being uniquely evaluated. In view of this, the current study aims to develop a method to take into account the shape of the gas cloud. The proposed method is designed to determine an equivalent gas cloud that can represent a specific number of actual clouds. The actual clouds refer to a group of gas clouds that are subdivided into the same category in the flammable gas cloud frequency distribution. Unlike the existing equivalent gas clouds, the proposed gas cloud has an inhomogeneous fuel concentration distribution whereby the shape can be implicitly considered. (C) 2019 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.